datadog-entity-generator
Fail
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: HIGHPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION] (HIGH): The skill is highly susceptible to Indirect Prompt Injection (Category 8) due to its core functionality of processing untrusted project data. * Ingestion points: The
scripts/project_analyzer.pyscript readsREADME.md,CODEOWNERS,pyproject.toml, and other source files which are attacker-controlled if the repository is untrusted. * Boundary markers: There are no boundary markers or instructions to ignore embedded commands mentioned in the workflow. * Capability inventory: The skill usesuv runto execute multiple local Python scripts and writes files to the.datadog/directory. * Sanitization: No sanitization or validation of the ingested metadata is described before it is used to generate YAML or drive the 'Engineer Interview'. - [COMMAND_EXECUTION] (MEDIUM): The skill performs broad filesystem scanning using
scripts/project_analyzer.py. The documentation explicitly lists.envas a target for metadata extraction. Reading environment files is a high-risk pattern (Category 2: Exposure) that can lead to credential leakage if the agent's output is redirected or if the logic is subverted via prompt injection.
Recommendations
- AI detected serious security threats
Audit Metadata